A fast modeling and conduction judgment method for three-dimensional random nanowire networks

By using kd-tree and parallel set algorithm in three-dimensional random nanowire networks and combining the reconstruction idea of ​​scapegoat tree, the problem of low efficiency in modeling and conduction judgment in the existing technology is solved, and the problem of low efficiency in modeling and conduction judgment in three-dimensional random nanowire networks is realized. Efficient conduction judgment and modeling simulation.

CN113536500BActive Publication Date: 2024-11-22SOUTHEAST UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202110783669.2
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-07-12
Publication Date
2024-11-22
Estimated Expiration
2041-07-12

AI Technical Summary

Technical Problem

In the prior art, the core algorithms are unclear and inefficient, especially the high complexity of conduction judgment, and the inability to effectively deal with extreme situations. The traditional kd-tree algorithm is deteriorated in extreme situations. .

Method used

The kd-tree algorithm is used to optimize the structure of the reconstruction idea of ​​the scapegoat tree, and the conduction situation is judged by using and checking algorithms. By randomly generating nanowires, judging through, inserting kd-tree, optimizing the tree structure and conduction judgment, Improve algorithm efficiency.

Benefits of technology

It realizes fast modeling and efficient conduction judgment of three-dimensional random nanowire networks, which is 5-10 times higher than traditional enumeration algorithms and 1.5 times higher than unmodified kd-tree algorithms, providing efficient solutions and experience.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN113536500B_ABST
    Figure CN113536500B_ABST
Patent Text Reader

Abstract

The present invention discloses a method for rapid modeling and conducting judgment of a three-dimensional random nanowire network, comprising the following steps: step 1, randomly generating a nanowire in a generation domain, and recording various coordinate parameters of the nanowire; step 2, using a kd-tree algorithm to find other nanowires adjacent to the nanowire generated in step 1, and judging whether they are interpenetrating, if interpenetrating, returning to step 1, if not interpenetrating, proceeding to step 3; step 3, inserting the nanowire generated in step 1 into a kd-tree storing all generated nanowires, and optimizing the kd-tree structure using the reconstruction idea of ​​a scapegoat tree; step 4, repeatedly executing steps 1-3 until the volume fraction of the generated nanowires reaches a predetermined requirement; step 5, using a union-find set to solve the conduction of the nanowire network. The present invention optimizes the efficiency of the modeling process, combines a union-find set algorithm, and more efficiently solves the conduction of the nanowire network.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of advanced nanomaterial design and analysis, and specifically relates to a method for rapid modeling and conducting judgment of a three-dimensional random nanowire network. Background Art

[0002] As a special form of solid material, random nanowire networks have become an important material foundation for emerging interdisciplinary disciplines such as modern electronics, microelectronics and materials science. Random nanowire networks composed of numerous metal nanowires have unique structures, excellent electrical properties, flexible mechanical responses, and adaptability to large-scale production. They have attracted much attention and have very broad application prospects in the fields of transparent conductive electrodes, flexible sensors, touch controls, displays, etc.

[0003] For random nanowire networks, calculating the performance parameters of the nanowire networks through computer simulation is an extremely effective means to pre-estimate the electrical and mechanical properties of the nanowire networks.

[0004] At present, the modeling and simulation of random nanowire networks are mostly based on two-dimensional random nanowire networks, while the simulation of three-dimensional random nanowire networks is relatively small. In addition, there are still the following deficiencies in the modeling and simulation of three-dimensional random nanowire networks:

[0005] For example, the core algorithm flow is not very clear, the algorithm process is not explained in detail, only the general principles are described, and there is a lack of analysis and testing of the algorithm's efficiency and operation.

[0006] For example, the algorithm for judging conduction is not efficient. The complexity of using the reachable matrix to judge in the three-dimensional generated domain is as high as O(N 3 )(N is the total number of nanowires in the generated domain), which is unacceptable in most cases.

[0007] For example, when using the mainstream kd-tree algorithm for modeling, if extreme generation situations are not specially considered, the algorithm efficiency may be degraded. Summary of the invention

[0008] The purpose of the present invention is to provide a method for rapid modeling and conduction judgment of a three-dimensional random nanowire network, so that the model can have good operating efficiency even for extreme data, and can efficiently perform conduction judgment of the nanowire network.

[0009] To achieve the above object, the technical solution adopted by the present invention is:

[0010] A method for rapid modeling and conducting judgment of a three-dimensional random nanowire network comprises the following steps:

[0011] Step 1, randomly generate a nanowire in the generation domain, and record various coordinate parameters of the nanowire;

[0012] Step 2, using the kd-tree algorithm to find other nanowires adjacent to the nanowire generated in step 1, and determine whether they penetrate each other. If so, return to step 1; if not, proceed to step 3;

[0013] Step 3, insert the nanowires generated in step 1 into the kd-tree storing all generated nanowires, and optimize the kd-tree structure using the reconstruction idea of ​​the scapegoat tree;

[0014] Step 4, repeatedly performing steps 1-3 until the volume fraction of the generated nanowires reaches a predetermined requirement;

[0015] Step 5: Use the union-find algorithm to solve the conductivity of the nanowire network.

[0016] Furthermore, in step 1, various coordinate parameters of the nanowire include: spatial coordinates of the two end points of the nanowire, spatial coordinates of the midpoint of the nanowire, serial number of the nanowire, and information of six planes of a bounding box surrounding the nanowire.

[0017] Furthermore, in step 2, the nanowires will be compared starting from the root node of the kd-tree. Each node of the kd-tree records the bounding box of the subtree with the node as the root. By judging whether the bounding box of the subtree intersects with the bounding box of the nanowire, it is pre-judged whether the nanowire has the possibility of intersecting with all the nanowires on the entire subtree.

[0018] Furthermore, in step 3, each node on the kd-tree records the number of nodes in the subtree with the node as the root. If it is found that the number of left subtree nodes and the number of right subtree nodes of a subtree are seriously unbalanced, the entire subtree will be reconstructed.

[0019] Furthermore, in step 5, the idea of ​​union-find is used to group all mutually conductive nanowires into the same set, and then the conductivity of the nanowires is obtained by judging whether a set contains nanowires connected to the upper and lower plates of the generating domain at the same time.

[0020] Beneficial effects: The present invention proposes a specific method for modeling and conducting three-dimensional random nanowire networks, which can improve the operating efficiency by 5-10 times compared with the traditional enumeration algorithm and about 1.5 times compared with the unmodified kd-tree algorithm; and the algorithm also provides a new idea for solving whether a random nanowire network is conducting, and has a substantial improvement in efficiency compared with the classic reachable matrix algorithm. Thus, it provides an efficient solution idea and experience for modeling and conducting three-dimensional random nanowire networks. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a schematic diagram of the flow of the three-dimensional random nanowire network modeling and the conduction judgment method of the present invention;

[0022] Figure 2 3D schematic diagram of the 3D random nanowire network model of the present invention; (the bold nanowires represent the conduction paths)

[0023] Figure 3 It is a schematic diagram of a certain conductive path of the three-dimensional random nanowire network of the present invention. DETAILED DESCRIPTION

[0024] The present invention will be further described below in conjunction with the accompanying drawings.

[0025] like Figure 1 The present invention provides a method for rapid modeling and conducting judgment of a three-dimensional random nanowire network, comprising:

[0026] Step 1: randomly generate a new nanowire in the generation domain and record various coordinate parameters of the nanowire:

[0027] All nanowires are generated in a L*D*H rectangular space, which is defined as the generation domain of the nanowire network.

[0028] In the simulation algorithm, each nanowire is stored as a record type, and the stored information includes:

[0029] a. Consider the nanowire as a line segment in three-dimensional space and record the spatial coordinates of the midpoint of the line segment (X mid ,Y mid ,Z mid );

[0030] b. Record the coordinates of the endpoints of the line segment (X1, Y1, Z1), (X2, Y2, Z2);

[0031] c. Record the number NUM of the nanowire;

[0032] d. Record the information of the nanowire bounding box: The coordinates of the two endpoints of the nanowire, a total of six components (X1, X2, Y1, Y2, Z1, Z2) are sorted into (X min ,X max ,Y min ,Y max ,Z min ,Z max ), then the bounding box can be represented by a six-tuple (X min -C,X max +C,Y min -C,Y max+C,Z min -C,Z max + C), where C is a constant representing the distance, which is generally half of the characteristic distance of the tunnel effect of the nanowire. min -C represents the plane X=X min -C, and similarly, the remaining five components have similar meanings, representing another five planes. These six planes form a cuboid that completely surrounds the nanowire, which is the bounding box of the nanowire.

[0033] Step 2: Use the kd-tree algorithm to find the nanowires adjacent to the new nanowire and determine whether they penetrate:

[0034] The nanowires will be compared starting from the root node of the kd-tree. The total bounding box of the entire subtree will be recorded on each root node of the kd-tree. By judging whether the bounding box of the subtree intersects with the bounding box of the nanowire, it can be judged in advance whether the nanowire has the possibility of intersecting with all the nanowires on the entire subtree.

[0035] The intersection of bounding boxes can be expressed using the following inequality:

[0036]

[0037] Among them, X min1 ,X max1 ,Y min1 ,Y max1 ,Z min1 ,Z max1 is the information of the first plane of the bounding box surrounding the nanowire, indicating that the plane X = X min1 ,X=X max1 ,Y=Y min1 ,Y=Y max1 ,Z=Z min1 ,Z=Z max1 The bounding box formed by min2 ,X max2 ,Y min2 ,Y max2 ,Z min2 ,Z max2 is the information of the second plane of the bounding box surrounding the nanowire, indicating that the plane X = X min2 ,X=X max2 ,Y=Y min2 ,Y=Y max2 ,Z=Z min2 ,Z=Z max2 The bounding box formed;

[0038] When the bounding boxes intersect, the shortest distance between the two line segments in space can be solved by mathematical methods, and the shortest distance between the two nanowires can be accurately solved. Let the shortest distance between the two nanowires be D, then the condition for the nanowire to penetrate is:

[0039] D<2r+t

[0040] Where r is the radius of the nanowire, and t is the characteristic distance of the tunnel effect. When the bounding box of the nanowire does not intersect with the bounding box of the subtree, it can be considered that the nanowire does not intersect with any nanowire in the subtree, and the judgment of the penetration of the nanowire with the nanowires in the subtree can be directly omitted, thereby improving efficiency. When the bounding box of the nanowire intersects with the bounding box of the subtree, first determine whether the nanowire and the nanowire represented by the root node penetrate each other. If penetration occurs, return directly to step 1. Then determine whether the bounding boxes of the left and right subtrees under the root node intersect with the bounding box of the nanowire. If intersection occurs, recurse to the left and right subtrees to perform the same judgment.

[0041] Step 3: Insert the new nanowire into the kd-tree that stores all generated nanowires, and optimize the kd-tree structure using the scapegoat tree reconstruction idea:

[0042] Each node on the kd-tree records the number of nodes in the subtree with the node as the root. If it is found that the number of nodes in the left subtree and the number of nodes in the right subtree of a subtree are seriously unbalanced, the entire subtree will be reconstructed as follows:

[0043] When the nanowire to be inserted is not found to intersect with other generated nanowires in step 2, it is considered that the generation position of the nanowire is in line with the actual situation, and the nanowire is inserted into the kd-tree. The insertion process is recursive, using the midpoint coordinates of the nanowire as the basis for judgment, and cyclically using the X, Y or Z dimensions at each layer. If the value of the nanowire to be inserted is less than the value of the root node in the judgment dimension of this layer, recurse to the left subtree of the tree, otherwise recurse to the right subtree of the tree until the nanowire to be inserted can be inserted into the kd-tree as a leaf node. When the insertion is completed, the bounding box information and size of each subtree are continuously updated during the backtracking process. When it is found that the size of the left subtree or right subtree of a subtree meets the size son >limit rebuild size tree When , reconstruct the subtree. son Indicates the size of the left or right subtree, limit rebuild is a constant coefficient representing the reconstruction threshold, size tree Indicates the size of the subtree.

[0044] Step 4, repeatedly perform steps 1-3 until the volume fraction of the generated nanowires reaches the predetermined requirement:

[0045] In order for the nanowire network to occupy a certain volume fraction in the generation domain, a number of nanowires need to be generated. Through steps 1 to 3, a nanowire can be generated. By repeatedly performing this process, a whole nanowire network can be generated.

[0046] Step 5: Use the union-find algorithm to solve the conductivity of the nanowire network:

[0047] By using the idea of ​​union-find, all interconnected nanowires are grouped into the same set. Then, by judging whether a set contains nanowires connected to the upper and lower plates of the generating domain at the same time, the conductivity of the nanowires is obtained. Specifically,

[0048] Nanowires can be divided into three categories according to their position in space: the first category is nanowires that intersect with the upper plate, the second category is nanowires that intersect with the lower plate, and the third category is nanowires that intersect neither the upper plate nor the lower plate. All nanowires have their own independent sets at the beginning. When two nanowires intersect, the algorithm will merge the sets to which the two nanowires belong. All nanowires in the same set are interconnected. When judging whether the entire nanowire network is conductive, it is only necessary to traverse all sets to determine whether both the first and second nanowires exist in the set.

[0049] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for rapid modeling and conducting judgment of a three-dimensional random nanowire network, characterized in that: The following steps are involved: Step 1, randomly generate a nanowire in the generation domain, and record various coordinate parameters of the nanowire; Step 2, using the kd-tree algorithm to find other nanowires adjacent to the nanowire generated in step 1, and determine whether they penetrate each other. If so, return to step 1; if not, proceed to step 3; Step 3, insert the nanowires generated in step 1 into the kd-tree storing all generated nanowires, and optimize the kd-tree structure using the reconstruction idea of ​​the scapegoat tree; Step 4, repeatedly performing steps 1-3 until the volume fraction of the generated nanowires reaches a predetermined requirement; Step 5, using the union-find method to solve the conductivity of the nanowire network; In step 2, the nanowires will be compared starting from the root node of the kd-tree. Each node of the kd-tree records the bounding box of the subtree with the node as the root. By judging whether the bounding box of the subtree intersects with the bounding box of the nanowire, it is pre-judged whether the nanowire has the possibility of intersecting with all the nanowires on the entire subtree. The intersection of the bounding boxes is expressed using the following inequality: Among them, X min1 ,X max1 ,Y min1 ,Y max1 ,Z min1 ,Z max1 is the information of the first plane of the bounding box surrounding the nanowire, indicating that the plane X = X min1 ,X=X max1 ,Y=Y min1 ,Y=Y max1 ,Z=Z min1 ,Z=Z max1 The bounding box formed by min2 ,X max2 ,Y min2 ,Y max2 ,Z min2 ,Z max2 is the information of the second plane of the bounding box surrounding the nanowire, indicating that the plane X = X min2 ,X=X max2 ,Y=Y min2 ,Y=Y max2 ,Z=Z min2 ,Z=Z max2 The bounding box formed; When the bounding boxes intersect, mathematical methods are used to solve the shortest distance between the two line segments in space and the shortest distance between the two nanowires. The shortest distance between the two nanowires is denoted as D. The condition for the nanowire to penetrate is: D<2r+t Where r is the radius of the nanowire, and t is the characteristic distance of the tunneling effect; When the bounding box of the nanowire does not intersect with the bounding box of the subtree, it is considered that the nanowire does not intersect with any nanowire in the subtree, and the judgment of the penetration of the nanowire with the nanowires in the subtree is directly omitted; when the nanowire intersects with the bounding box of the subtree, first determine whether the nanowire penetrates the nanowire represented by the root node. If penetration occurs, return to step 1 directly, and then determine whether the bounding boxes of the left and right subtrees under the root node intersect with the nanowire bounding box. If intersection occurs, recurse to the left and right subtrees to perform the same judgment.

2. The method for rapid modeling and conducting judgment of a three-dimensional random nanowire network according to claim 1, characterized in that: In step 1, various coordinate parameters of the nanowire include: spatial coordinates of the two end points of the nanowire, spatial coordinates of the midpoint of the nanowire, serial number of the nanowire, and information of six planes of a bounding box surrounding the nanowire.

3. The method for rapid modeling and conducting judgment of a three-dimensional random nanowire network according to claim 2, characterized in that: The information of the six planes of the bounding box surrounding the nanowire is: the coordinates of the two end points of the nanowire, a total of six components (X1, X2, Y1, Y2, Z1, Z2) are sorted into (X min ,X max ,Y min ,Y max ,Z min ,Z max ), then the bounding box can be represented by a six-tuple (X min -C,X max +C,Y min -C,Y max +C,Z min -C,Z max +C), where C is a constant representing the distance, which is half of the characteristic distance of the tunnel effect of the nanowire; X min -C represents the plane X=X min -C, similarly, the remaining five components have similar meanings, representing another five planes. These six planes form a cuboid that completely surrounds the nanowire, which is the bounding box of the nanowire.

4. The method for rapid modeling and conducting judgment of a three-dimensional random nanowire network according to claim 1, characterized in that: In step 3, each node on the kd-tree records the number of nodes in the subtree with the node as the root. If it is found that the number of nodes in the left subtree and the number of nodes in the right subtree of a subtree are seriously unbalanced, the entire subtree will be reconstructed.

5. The method for rapid modeling and conducting judgment of a three-dimensional random nanowire network according to claim 4, characterized in that: In step 3, when the nanowire to be inserted is not found to intersect with other generated nanowires in step 2, it is considered that the generation position of the nanowire is in line with the actual situation, and the nanowire is inserted into the kd-tree; the insertion process is recursive, using the midpoint coordinates of the nanowire as the basis for judgment, and cyclically using the X, Y or Z dimensions at each layer. If the value of the nanowire to be inserted is less than the value of the root node in the judgment dimension of the layer, recursively go to the left subtree of the tree, otherwise recursively go to the right subtree of the tree, until the nanowire to be inserted can be inserted into the kd-tree as a leaf node; When the insertion is completed, the bounding box information and the size of each subtree are continuously updated during the backtracking process; when it is found that the size of the left subtree or the right subtree of a subtree meets the size son >limit rebuild size tree When , reconstruct the subtree; where size son Indicates the size of the left or right subtree, limit rebuild is a constant coefficient representing the reconstruction threshold, size tree Indicates the size of the subtree.

6. The method for rapid modeling and conducting judgment of a three-dimensional random nanowire network according to claim 1, characterized in that: In step 5, all mutually conductive nanowires are grouped into the same set by using the idea of ​​union-find, and then the conductivity of the nanowires is obtained by judging whether a set contains nanowires connected to the upper and lower plates of the generation domain at the same time.

7. The method for rapid modeling and conducting judgment of a three-dimensional random nanowire network according to claim 6, characterized in that: In step 5, the nanowires are divided into three categories according to their positions in space: the first category is the nanowires intersecting with the upper plate, the second category is the nanowires intersecting with the lower plate, and the third category is the nanowires intersecting with neither the upper plate nor the lower plate; all the nanowires initially have their own independent sets, and when two nanowires intersect, the sets to which the two nanowires belong are merged, and all the nanowires in the same set are interconnected; when determining whether the entire nanowire network is conductive, it is only necessary to traverse all the sets to determine whether both the first category of nanowires and the second category of nanowires exist in the set.